Identification of objects in an image captured by a digital imaging device may rely on extracting features of the object and use pattern recognition of those features to identify the object. Pattern recognition and feature extraction may be subject to inaccuracies caused by, among other things, lighting conditions, changes in object orientation, relative distance of the object from the imager and occlusion. Typical identification processes may therefore entail rotation, translation and scale invariant features, and may call for complicated pattern recognition algorithms. Adding an object to the recognition task may therefore require adjusting a recognition algorithm compensation for lighting condition.